\begin{table}[H] \centering
\newcolumntype{R}{>{\raggedleft\arraybackslash}X}
\newcolumntype{L}{>{\raggedright\arraybackslash}X}
\newcolumntype{C}{>{\centering\arraybackslash}X}

\caption{Heterogeneity with respect to government experience of challengers}
\label{tab:quality_bounding}
\begin{tabularx}{\linewidth}{@{}lCCC@{}}

\toprule
 & & \multicolumn{2}{c}{\textbf{Keeping challengers that have}} \tabularnewline & (1) & (2) & (3)  \tabularnewline 
{}&{Baseline}&{Not been prev. in power}&{Been prev. in power} \tabularnewline
\midrule \addlinespace[\belowrulesep]
\textbf{Economic performance}&0.269***&0.243**&0.286** \tabularnewline
&(0.101)&(0.117)&(0.149) \tabularnewline
\textbf{HDI}&0.197&0.246&0.065 \tabularnewline
&(0.167)&(0.211)&(0.266) \tabularnewline
\textbf{Democracy}&0.192**&0.118&0.437** \tabularnewline
&(0.101)&(0.113)&(0.213) \tabularnewline
\textbf{General index}&0.276***&0.226**&0.290** \tabularnewline
&(0.105)&(0.116)&(0.142) \tabularnewline
\bottomrule \addlinespace[\belowrulesep]

\end{tabularx}
\\ \parbox{\linewidth}{\footnotesize \caption*{\footnotesize \emph{Notes}: This table presents the results of an exercise in which we restrict the sample to elections in which the challenger has not (resp., has) won elections in the past decade and was plausibly of lower (resp., higher) quality (in column 2, resp. 3). $^{*} p<0.10,^{**} p<0.05,^{***} p<0.01$.}}
\end{table}
